Co-breeding Association of Aedes albopictus (Skuse) and Aedes aegypti (Linnaeus) (Diptera: Culicidae) in Relation to Location and Container Size

The occurrence of major outbreaks of dengue, and other vector borne diseases such as chikungunya and zika in tropical and subtropical regions has rendered control of the diseases a top-priority for many affected countries including Malaysia. Control of the mosquito vectors Aedes aegypti and Aedes albopictus through the reduction of breeding sites and the application of insecticides to kill immature forms and adults are the main control efforts to combat these diseases. The present study describes the association between Ae. albopictus and Ae. aegypti in shared breeding sites. This study is important given that any measure taken against one species may affect the other. A yearlong larval survey was conducted in four dengue endemic areas of Penang Island. Sorenson’s coefficient index indicated that no association between number of the immatures of the two species regardless of container size and study location. Therefore, the mean number Ae. albopictus immature was not decreased in the presence of Ae. aegypti in shared breeding container. However Ae. aegypti appeared to prefer breeding in habitats not occupied by Ae. albopictus, the two species sharing breeding sites only where available containers were limited. In control efforts, eliminating the preferred breeding containers for one species might not affect or reduce the population of the other species.


INTRODUCTION
A species would not be able to survive on its own but lives together with other organisms to form a community in the same habitat. Co-existence of multiple species of mosquitoes in a habitat at a given time indicates positive interaction among them (Pemola Devi & Jauhari 2007). Interspecific associations among mosquitoes are often related to physicochemical and biological composition of mosquito breeding waters (Reisen et al. 1981;Almiron & Brewer 1996;Rajnikant et al. 1998). Often, interspecific association shows similarity of habitat requirements and interactions between species (Cole 1949). However, past research on the breeding of different species under field conditions have been based mainly on the frequency of co-occurrence of the immature stages without rigorous statistical analysis to validate the strength of association or repulsion between them (Cole 1949;Bhat 1975a;Bhat 1975b;Malhotra et al. 1987;Bhat et al. 1990).
Aedes albopictus, is believed to have originated from the tropical forests of Southeast Asia (Smith 1956). Meanwhile, Ae. aegypti originated from Africa (Mousson et al. 2005). In Malaysia, the first occurrence of Ae. aegypti was recorded by Leicester in 1908 and Stanten in 1914 (Lee & Cheong 1987). Population of Aedes albopictus and Ae. aegypti are common in urban and rural areas of Malaysia (Nazni et al. 2009;Saifur et al. 2013;Basari et al. 2016). Both are sympatric species, and coexist in similar habitat (Klowden 1993;Gilotra et al. 1967;Sprenger & Wuithirsnysgool 1986;O' Meara et al. 1993;Chen et al. 2006a). Whenever, two species try to coexist in same ecological niches, species replacement or displacement tend to occur. Hawley (1988) reported that species replacement occurred in particular in the North of America, where Ae. aegypti abundance had been reduced as a result of competition with Ae. albopictus.
Whereas, Juliano (1998) reported displacement of Ae. aegypti by Ae. albopictus which might due to larval competition on available food resource. Others reported replacement of Ae. albopictus by Ae. aegypti in the peripheral areas of towns of India (Kalra et al. 1997).
On the contrary, others reported Ae. aegypti has completely replaced the indigenous Ae. albopictus in urban areas (Pant et al. 1973;Service 1992). In Bangkok, Thailand and Calcutta, India, Ae. albopictus had decreased in population, while Ae. aegypti has become more pronounced (Rudnick & Hammon 1960;Gilotra et al. 1967). In addition, experiments conducted in controlled environment support the proposition that Ae. aegypti can out-compete and displace Ae. albopictus (Moore & Fisher 1969;Sucharit et al. 1978;Black et al. 1989). Lambrechts et al. (2010) hypotesize, Ae. aegypti is gradually replacing Ae. albopictus as the dominant day-biting mosquito in Asian cities because it is better adapted to the urban environment.
In Penang Island, Malaysia, during the year 1970's, Ae. aegypti was not documented beyond the city limit of Georgetown to the rest of the island (Yap 1975). However, recent studies, shows that the species is observed in rural and urban residential areas of the island (Saifur et al. 2013). Compared with field study done in Northern Queensland, Australia, it was noted that Ae. notoscriptus container distribution was affected by the presence or absence of Ae. aegypti although they found no association in the relative abundance of both species (Tun Lin et al. 1999). In summary, researchers noted Ae. aegypti and Ae. albopictus do mix-breed in large-sized water containers regardless whether it is an indoor or outdoor environments (Hwang & Hsu 1994;Chen et al. 2006a).
Therefore, for an effective mosquito control regime, the relationship between habitats, environmental factors and occurrence of immature mosquitoes must be well understood. The association between species of mosquitoes can provide clues to better understanding of their biology and roles in the transmission of the vector borne viruses such as dengue. Therefore, it is important to determine the strength of association of these two species, in respect of positive association (overlapping), negative association (repulsion) or zero association (the species is independent). This study sought to determine if there was co-breeding association between Ae. aegypti and Ae. albopictus in shared breeding containers in the Southwest district of Penang, Malaysia.

Study Sites
Four areas within Southwest district of Penang Island were selected for this study (Fig. 1). The areas were: Pantai Jerjak (urban residential area) located at 5.337681° N, 100.302187° E (12 m.a.s.l.), Bayan Lepas (urban residential/industrial area) located at 5.298113° N, 100.262276° E (14 m.a.s.l.), Batu Maung (suburban residential area) located at 5.274604° N, 100.267525° E (8 m.a.s.l.) and Balik Pulau (rural area) located at 5.325033° N, 100.212108° E (18 m.a.s.l). Climatological data for Penang Island including rainfall, mean relative humidity and mean temperature were obtained from the Malaysian Meteorological Station located at Penang International Airport.

Larval Survey
Larvae collection was done for 12 months (January 2009 to December 2009). The sampling were performed on monthly basis in each study area stated above by three two-person collection teams (between 0900 h and 1500 h). A total of 720 houses in each study area were inspected for mosquito breeding sites. During sampling, an inspection of the domestic and peri-domestic area of each house in study areas for water holding containers was performed. The containers were categorised into three sizes: small (capacity < 1 litre), medium (1 litre < capacity < 15 litres) and large (capacity > 15 litres). Due to the different sizes of the containers, sampling methods for the three container categories also differed. For small containers, all the contents of the containers were poured into zip-lock plastic bags, while for medium and large containers only the Aedes immatures (pupae and larvae) were collected using pipette or sieves and placed into zip lock plastic bags. All the water samples (in plastic bags) were labelled with house description and container name so that samples could be linked to the exact container and household of origin. The containers with mosquitoe's larvae were also classified into three categories: i. Single container -with either Ae. albopictus or Ae. aegypti ii. Shared container -Ae. albopictus and Ae. aegypti together iii. Other container -other mosquito species, Aedes absent The samples were transported to the laboratory at the School of Biological Sciences, Universiti Sains Malaysia on the same day they were collected for further processing. For the purpose of identification, pupae were reared and identified when they developed into adults. The 1st and 2nd instar larvae were allowed to moult to the 3rd and 4th instar to facilitate identification; 3rd and 4th instar were identified to the species level using taxonomic keys provided by Rueda (2004) under a dissecting microscope (Olympus CX41, Olympus, Tokyo, Japan).

Data Analyses
The Chi-square test was used to analyse differences in container abundance, immature abundance, immature species and container sizes using the SPSS version 21.0.

Coefficient of Interspecific Association
The method of Fager (1957) as detailed by Southwood (1978) was used to explain the independence or association of the two species. An association between Ae. albopictus and Ae. aegypti would be determined from the proportion of positive containers containing Ae. albopictus in the presence or absence of Ae. aegypti in the same containers. If there is no association, the same proportions of Ae. aegypti should be observed irrespective of whether Ae. albopictus was present or not (Tun Lin et al. 1999).
To calculate the coefficient of association, 2 × 2 contingency tables were drawn up where a, b, c and d were the number of occurrences of the two species in water containers as shown in the  (1977): Where, m = (a+b), n = (c + d), r = (a + c), s = (b + d) and N = m + n + r +s

Index of Association (I)
The proportion of individuals occurring together was calculated using Sorensen's Coefficient Index (1948) as modified by Southwood (1978). The formula was as follows: where J = the number of Ae. albopictus and Ae. aegypti immatures where the two species shared positive containers, A = the number of Ae. albopictus immature found in all positive containers B = the number of Ae. aegypti found in all positive containers.
An Index value of +1 indicates complete association while −1 indicates no association.

Dominance Index (D)
Species dominance, D, using May's (1975) index was calculated for each study site and container size: Where Y max = the number of larvae of the most common species (Ae. albopictus) in the each study site or each container size, Y t = the total numbers of larvae of all species in the habitat.

RESULTS
During the larval survey, the monthly mean temperature and mean relative humidity in Penang Island ranged between 26.0°C to 28.0°C and 59% to 89% respectively. Overall, Penang received a total rainfall of 2407.6 mm.   Note; *significant, p < 0.05, Where, a = the presence of both species (A and B) in shared containers, b = the presence of species A but species B absent, c = the presence of species B but species A absent, d = samples of other Aedes species but species A and B absent, bc = single containers for both species, ad = both species in shared containers and negative containers, positive association when ad-bc = +ve, negative association/repulsion when ad-bc = −ve.
However, when comparison were made, in terms of the abundance of immatures, there was no co-breeding association between Ae. albopictus and Ae. aegypti in each study area (Table 3). Similarly, the analysis between Ae. albopictus and Ae. aegypti immature relative abundance in three different sizes of containers indicated no co-breeding association between both species in each container size (Table 4).  Tables 5 and 6 shows the species dominance index calculated for each study area and container size, respectively. The results showed that Ae. albopictus was the dominant species (> 90%) for all study areas. Thus, Ae. albopictus is the dominant Aedes species in the Southwest district of Penang Island, regardless whether it is urban, suburban or rural area.  Where Y max = the number of immatures of the most common species (Ae. albopictus) in each survey areas, Y t = the total number of immatures of all species in the areas.

DISCUSSION
According to Hurlbert (1969), the analysis of presence-absence data is preferable to that of the relative number of immature stages for measuring the degree of association between two species. However, Southwood (1978) suggested to employ both methods. Positive association means two species interact in such a way as to favour mutual presence. Negative association is to be anticipated when one species exclude the other from the habitat. In the present study, Ae. albopictus and Ae. aegypti were found in single and shared containers regardless of the geographical characteristics (urban, suburban) and container size (small, medium, large). Negative value for (ad-bc) (Table 2) indicated that there was a negative association between the two species which indicated that Ae. aegypti preferred to fill in habitats which were not occupied by Ae. albopictus. It is possible that after entering houses to blood-feed, Ae. albopictus found indoor containers which had been occupied by Ae. aegypti when water holding containers outdoor dried out during the dry season. Past research observed Ae. albopictus do oviposit indoors in human dwellings (Sulaiman et al. 1991;Chen et al. 2006b;Lian et al. 2006;Wan-Norafikah et al. 2010;Dieng et al. 2010) and the most anthropophilic mosquito in Malaysia (Parker et al. 1983).
Negative co-breeding association between the two species in all container sizes confirmed that Ae. albopictus would fill out niches unoccupied by Ae. aegypti. The latter prefers to breed in both indoor and outdoor containers where vegetation in the areas are less. Previous studies showed Ae. aegypti to be the dominant indoor species (Surendran et al. 2007;Singh et al. 2008;Wan-Norafikah et al. 2010). According to Gilotra et al. (1967), Ae. aegypti is the superior competitor in domestic premises, whereas Ae. albopictus has the advantage in outdoor or silvatic surroundings.
The Sorenson's Coefficient Index showed there was no significant association between individual immature for the two species and no association in relative abundance between individual species. Similar results were obtained in all the study areas and each container size. The index value for urban and suburban was similar suggesting dominance of Ae. albopictus in small and medium size containers. The mean immature densities of Ae. albopictus were not depressed significantly in the presence of Ae. aegypti. Aedes albopictus continued to be the dominant Aedes species in the Southwest district of Penang Island despite the spread of Ae. aegypti out of the city limit. Similarly, Tun-Lin et al. (1999) found that there was a significant co-breeding association in the distribution of positive containers for Ae. notocriptus depending on the presence and the absence of Ae. aegypti in Australia. They also found that there was little or no association between the two species in their relative abundance of immatures in shared containers.
Being the dominant Aedes species in the Southwest district of Penang Island, Ae. albopictus might play an important role in the transmission of dengue and chikungunya viruses. According to Lounibos (2002), though Ae. aegypti is the main dengue vector, Ae. albopictus is also a competent vector and may be locally important. Interspecies competition between larvae change Ae. albopictus behaviour. In shared breeding habitat, the larvae adapted by swimming faster, increased their movement and feeding rate. Breeding containers with high larvae density tend to have limited space and resource. Therefore, Ae. albopictus larvae that have less food during development will emerge as an adult smaller in size which was reported to affect its fitness, reproductive rate and capacity as a vector (Blaustein et al. 2005;Preisser et al. 2005;Bara et al. 2015).
For Ae. albopictus, competition increases the probability of obtaining arboviruses (Alto et al. 2005;Alto et al. 2008) and competition among larvae may affect the probability of vector-borne virus transmission (Alto et al. 2008). Furthermore, effects of competitive interactions among larval stages may be carried over to the adult stage and affect vector competence, which describes the ability to become infected and subsequently to transmit a pathogen after imbibing an infectious blood meal (Hardy 1988).
When comparing larval competition in co-exist populations, Ae. aegypti stand a better chance as it requires shorter developmental time than Ae. albopictus (Chan et al. 1971). However, Phon (2007) indicated the competitive advantage of Ae. albopictus in situations of limited resources could be the reason for the dominancy of this mosquito in Penang Island. Barrera (1996) noted the presence of rapidly decaying detritus (e.g., animal detritus) tends to yield competitive equality or advantage for Ae. aegypti, whereas refractory plant detritus (deciduous or coniferous leaves) tends to yield competitive advantage for Ae. albopictus. He also emphasized the interspecific differences in starvation resistance of larvae of these species also depended on type of food resource. Aedes albopictus and Ae. aegypti were found to withstand starvation when reared on oak leaves and liver powder, respectively, suggesting a physiological basis for the detritus-typedependence having an impact on co-breeding competition of these two species.
In Australia, Tun-Lin et al. (1999) proposed the association between Ae. aegypti and Ae. notoscriptus could be due to competitive displacement of immature stages, different adult ovipositional stimuli or pheromonal repellents. However, competitive displacement of Ae. albopictus by Ae. aegypti in Penang Island is unlikely to happen. Shared breeding between Ae. albopictus and Ae. aegypti encountered in the present study was very low. The present study demonstrated that there was negative co-breeding association between the two species in their container distribution (number of container) and no association existed between the number of immatures of both species. Therefore, statistically, the interaction was significant only in the number of containers occupied by both species but there was no interspecies association from the perspective of individual mosquitoes.
The spread of Ae. aegypti in Penang Island could be due to several factors such as the rapid and extensive urbanisation of the city, the difference in fecundity between Ae. aegypti and Ae. albopictus and the difference in the duration of the life cycle of the two species. Favourable environment for the highly domesticated Ae. aegypti has been created with rapid and extensive urbanisation, and this condition leading to the rapid spread and increase in numbers of the species. However, Ae. albopictus probably has never been displaced by Ae. aegypti from the urban areas since the current trend of urban development is towards a 'garden city' where habitats would still be available for Ae. albopictus (Chan et al. 1971).

CONCLUSION
Negative interspecific association was observed between Ae. albopictus and Ae. aegypti in breeding containers in four survey areas of Southwest district on Penang Island suggesting Ae. aegypti distribution is restricted by Ae. albopictus. In addition, though Ae. albopictus and Ae. aegypti share the same breeding habitat, both prefer different environments (indoor or outdoor). The two species would avoid breeding in the same containers. Therefore, as the two species have different preferences in the selection of breeding environment, mosquito control should be emphasised in both inside and outside areas.